Cisco’s Journey to Predictive Services: Three Lessons from Hollywood
“It took me 17 years and 114 days to become an overnight success.” – Lionel Messi
This week, we launched Cisco Predictive Services to help customers respond to the need for speed, fill the IT talent gap, and grow their business. These services leverage Machine Learning (ML) to constantly adapt, learn and protect so that CIOs can unlock growth and mitigate risk.
74% of IT executives expect IT services to predict potential IT issues and take action to avoid them. Our two new Cisco Services portfolios, Business Critical Services and High-value Services usher a new era of predictive services:
- At a US bank, we used ML to predict an outage 32 hours before it would have occurred.
- At a service provider, we automated 99.9% of software changes – raising CSAT and lowering outages.
Cisco has been investing for 30 years to make this new era of services a reality. In the last 2-3 years, we doubled down on analytics and ML. The journey has not been easy and there is still a lot more to do. We have learned a ton in the process. As I reflect on our journey, three lessons stand out.
I love movies. So you will find the lessons expressed in Hollywood metaphors.
#1. The Iron Man Lesson
Be explicit about where you want to be on Man-Machine continuum.
Some believe machines will replace humans, and others believe machines will merely supplement humans. After thoughtful debate, we’ve concluded that for our services business it’s not a binary choice, but rather a conscious decision on where we want to be on the Man-Machine continuum. Our people, who are expert advisors, simplifiers, implementers and optimizers, are trusted partners of our customers. We believe that the human connection and human judgement will always remain essential.
Consider “Iron Man” – a 2008 American superhero film based on the Marvel Comics character. The story is about Tony Stark, an engineer who builds a powered exoskeleton. While Tony Stark and the exoskeleton are both great, the combination of the two makes Iron Man the awesome super-hero my six-year old son loves!
So, we took a lesson from Iron Man. In the example of the bank, ML generates the good/bad patterns, and a Cisco Services expert reviews them and makes edits. Cisco software recommends steps to auto-remediate problems, and our human experts vet them before making changes.
#2. The Moneyball Lesson
Take a data-backed approach to scouting and nurturing ML initiatives.
There is a lot of hype around ML. Instead of starting with real problems that need to be solved and then using ML as a tool, it sometimes feels like people use it as a hammer and then look for nails. So, how do you find solutions that use ML to deliver true differentiated value to customers?
I like to think of this problem through the lens of the movie “Moneyball” – a 2011 American film based on Michael Lewis’s 2003 nonfiction book. While most baseball scouts rely on qualitative metrics and hunches to select “stars,” in this movie Billy Beane (the general manager of the Oakland Athletics team) uses data to find and nurture overlooked talent, enabling him to assemble an unexpected winning team.
At Cisco Services, innovation happens at all levels, and some of our best innovation happens at the grassroots level. A small group of engineers started digitizing Cisco’s intellectual capital and created a solution that led to an exponential reduction in time to resolution of cases. They called the solution, “Big Data Broker (or BDB) BORG.”
Today, BDB BORG can process customer files against 17,000+ modules/scripts. In these scans, we detect at least one problem 86% of the time, and 46% of the time BORG detects a high-severity issue that should be resolved right away. BDB BORG, with its humble beginnings, won Cisco’s Pioneer Award – the highest accolade for innovation within Cisco.
#3 The Minority Report Lesson
Be relentless about building “the” data-analytics platform.
Cisco did not start its Predictive Services efforts greenfield. There were ongoing data/analytics efforts in various parts of our business, and we decided to pull our various data sets together into one platform.
This reminds me of “Minority Report” – a 2002 American science fiction film directed by Steven Spielberg, loosely based on the 1956 short story by Philip K. Dick. In the movie, “PreCrime,” a specialized police department, apprehends criminals before they commit their crimes, based on data provided by three psychics. While each of the psychics have powers, it is the combination of the three that leads to the best results.
Cisco realized that it must tap the combined power of its data sets. We have installed 50M+ networks over the last 20 years, and we conduct 6M+ customer support interactions every year. We have billions of pieces of digitized knowledge. With the help of our Distinguished Engineer community and a central analytics group, we put together a proposal for a single platform. At a service provider, this analytics platform powered by multiple data sources led to a 30% reduction in time-to-resolution.
I am proud to have been part of the many teams that have worked tirelessly over many months to bring this next generation of Cisco Services to life. We will continue to learn, lead and innovate, because we are committed to helping our customers accelerate their digital transformations.